A range of behavioral, physiological, and cognitive responses (e.g. approach and avoidance, autonomic reactivity, and subjective feelings) reflects a subject's emotional state. The cognitive regulation of emotion refers to the capacity to regulate these emotional responses in a flexible manner according to a cognitive operation. Deficits in the cognitive regulation of emotional processes characterize many psychiatric disorders. In everyday life, however, particular sensory stimuli and/or actions can elicit different emotional responses depending upon the situation or context. Contexts often rely on a cognitive understanding of one's current situation in the absence of explicit cues. These types of contexts may be referred to as ?abstract? contexts. This grant studies a type of abstract context where the context is determined by a task set. A task set is the set of stimulus- response-outcome mappings (or rules) that dictate correct performance for trials within a particular block. Previous research demonstrates the capacity of primates to learn these abstract contexts, and neural representations of abstract contexts exist in the amygdala and two areas in the prefrontal cortex (PFC), the anterior cingulate and orbitofrontal cortices (ACC and OFC). This grant seeks to understand the mechanisms that underlie the formation and maintenance of these representations of contexts. In contrast to supervised learning driven by error signals, we hypothesize that the occurrence of temporally associated trial types triggers unsupervised learning, presumably through a Hebbian mechanism involving activity-dependent plasticity. This learning could underlie formation of representations of abstract contexts defined by task sets, which will be explored with electrophysiological recordings in Aim 1. The creation of a representation of a task set requires combining information about the current trial with information about the trials that have occurred recently. Brain structures that provide memory traces of recent events and/or that combine information over time could create representations of a task set prior to the emergence of the representations observed in amygdala, OFC, and ACC. Our next experiments therefore target the hippocampus and dorsolateral PFC (DLPFC), which are implicated in memory processes, working memory, and executive functions. We will compare and contrast the encoding of task sets in hippocampus, DLPFC, OFC, and ACC during and after learning about task sets (Aim 2). Finally, we will use causal methods to determine if PFC input to the amygdala and the hippocampus acts to maintain these context representations, which could be a vital mechanism for the cognitive regulation of emotion (Aim 3). Overall, these experiments promise to illuminate neurophysiological mechanisms critical for normal adaptive emotional health. !